Skogby et al.

1

Cover page 1

Water incorporation in omphacite: concentrations and compositional 2

relations in ultrahigh-pressure eclogites from Pohorje, Eastern Alps 3

4

Running title: Water incorporation in omphacite 5

Plan: 6

Abstract 7

Introduction 8

Geological background and petrography 9

Experimental methods 10

Infrared spectroscopy 11

Mössbauer spectroscopy 12

Electron microprobe analysis 13

Results 14

Discussion 15

Acknowledgements 16

References 17

18

Corresponding author: 19

Henrik Skogby 20

Swedish Museum of Natural History 21

Department of Geosciences 22

Box 50007 23

SE-104 05 Stockholm 24

Sweden 25

e-mail: henrik.skogby@nrm.se. 26

phone: +46 8 51954043 27

28

29

30

mailto:henrik.skogby@nrm.se

Skogby et al.

2

31

Title page 32

33

Water incorporation in omphacite: concentrations and compositional 34

relations in ultrahigh-pressure eclogites from Pohorje, Eastern Alps 35

36

Henrik Skogby1*

, Marian Janák2, and Igor Broska

2 37

38 1Swedish Museum of Natural History, Department of Geosciences, Box 50007, SE-104 05 39

Stockholm, Sweden. 40

* Corresponding author, e-mail: henrik.skogby@nrm.se. 41 2Earth Science Institute, Slovak Academy of Sciences, Dúbravská 9, Box 106, 840 05 42

Bratislava, Slovak Republic. 43

44

45

Skogby et al.

3

46

Abstract: Omphacite in ultra-high pressure (UHP) eclogites from the Pohorje Mountains in 47

Slovenia, south-eastern Alps, has been investigated by electron microprobe (EMP), Infra-Red 48

(IR) and Mössbauer spectroscopy to determine OH concentrations and related incorporation 49

mechanisms. Results from polarized IR measurements reveal high contents of structurally-50

bound OH, varying from 530 to 870 ppm H2O. Characterization of omphacite composition by 51

EMP analysis and Mössbauer spectroscopy shows that all samples contain vacancies at the 52

M2 position, which can be expressed as a Ca-Eskola component (Ca0.5□0.5AlSi2O6). The 53

amount of the Ca-Eskola component displays a positive correlation with the OH 54

concentration, which confirms results from previous studies. The occurrence of precipitated 55

quartz rods in some samples indicates that primary omphacite contained a larger Ca-Eskola 56

component. Extrapolation of the observed trend-line for the Ca-Eskola and OH contents point 57

to an original OH concentration around 1500 ppm H2O for these samples. The high water 58

contents observed in omphacite are considered to be linked to the UHP origin of the eclogite 59

rocks. 60

61

Key words: Omphacite; Ca-Eskola; eclogite; UHP; infrared spectroscopy; Mössbauer 62

spectroscopy. 63

64

65

Introduction 66

67

The pyroxenes are one of the nominally anhydrous mineral groups that have been shown to 68

generally contain low but significant contents of hydrogen, bound in their crystal structures as 69

OH- ions. The amounts of hydrogen incorporated in pyroxene are considered to be related to 70

both intrinsic factors, such as sample composition and defect chemistry, as well as to external 71

Skogby et al.

4

ones such as T, Ptot and PH2O conditions, thereby reflecting the equilibration environment. 72

During recent years, hydrogen contents in pyroxenes have been extensively explored to gain 73

information from a range of geological environments, including water storage mechanisms 74

and circulation in Earth’s mantle (e.g. Ingrin & Skogby, 2000; Bolfan-Casanova, 2005; 75

Peslier, 2010) as well as magmatic water contents in volcanic systems (e.g. Wade et al., 2008; 76

O’Leary et al., 2010; Weis et al., 2015). 77

The modes of hydrogen accommodation in clinopyroxene are not yet fully understood. 78

However, most incorporation models (e.g. Andrut et al., 2007) have suggested that OH 79

preferentially occurs on the underbonded O2 position, with charge compensation provided 80

either by vacancies at the M2 position or charge-deficient substitutions such as Al replacing 81

Si at the tetrahedral site. Among the pyroxenes, omphacite has been shown to frequently 82

accommodate relatively high concentrations of OH (Skogby et al., 1990, Smyth et al., 1991; 83

Katayama & Nakashima, 2003; Katayama et al., 2006; Koch-Müller et al., 2004; 2007; 84

Konzett et al., 2008a), approaching close to 1000 ppm H2O. These high contents appear to be 85

coupled to formation of structural vacancies via the so-called Ca-Eskola (CaEs) component, 86

Ca0.5□0.5AlSi2O6, which has been shown to be essential for OH incorporation in omphacite. 87

Such a relation was first reported by Smyth et al. (1991) for a series of jadeite-rich 88

omphacites from South-African kimberlites, where the amount of M2 vacancies was shown to 89

be correlated with the absorbance of the OH-bands in FTIR spectra, and hence water 90

concentration. This trend was further supported by Katayama & Nakashima (2003), who also 91

found a positive correlation for OH absorbance and calculated CaEs components in a series of 92

clinopyroxenes from deeply subducted crust occurrences in the Kokchetav UHP metamorphic 93

terrane. In an experimental study of hydroxyl solubility in synthetic jadeite and Na-rich 94

clinopyroxene, Bromiley & Keppler (2004) concluded that solid solutions of jadeite with 95

diopside and in particular CaEs components lead to a drastic increase of water solubility, and 96

Skogby et al.

5

that these compositional effects were substantially larger than those imposed by pressure and 97

temperature. Additional support for the importance of the CaEs component was provided by 98

Koch-Müller et al. (2004) for a series of omphacites from Yakutian kimberlite pipes. 99

Even though the studies mentioned above have clearly demonstrated that the OH 100

content in omphacite is frequently correlated with the CaEs component, a full characterization 101

of the amount of CaEs has generally not been established. This has been caused by lack of Fe 102

valence determination, which is needed to calculate an appropriate CaEs (and vacancy) 103

component, and has lead to underestimation and often negative values of the CaEs component 104

when all Fe has been assumed as Fe2+

. In order to increase the understanding of OH 105

incorporation in omphacite and the role of the CaEs component, we have here investigated a 106

series of omphacites from UHP eclogites by FTIR spectroscopy and microprobe analysis, as 107

well as Mössbauer spectroscopy to fully characterize the cation contents of the structural 108

formulae. 109

110

Geological background and petrography 111

112

The investigated eclogites come from the Pohorje Mountains in Slovenia, at the south-eastern 113

margin of the Alps (Fig. 1). The Pohorje Mountains is mainly formed by metamorphic rocks 114

of continental basement type, predominantly micaschists, gneisses, marbles, and 115

metaquarzites (Mioč & Žnidarčič, 1977; Kirst et al., 2010), intruded by granodioritic to 116

tonalitic plutons and pegmatites in the Miocene (Altherr et al., 1995; Fodor et al., 2008; 117

Trajanova et al., 2008; Uher et al., 2014). Eclogites, partly amphibolitised, occur in the south-118

eastern part of Pohorje, together with meta-ultramafic rocks of the Slovenska Bistrica 119

Ultramafic Complex (SBUC; Hinterlechner-Ravnik et al., 1991; Janák et al., 2006; De Hoog 120

et al., 2009), metasedimentary gneisses and marbles. 121

Skogby et al.

6

The first evidence for UHP conditions was documented in kyanite-bearing eclogites 122

(Janák et al., 2004). Subsequently, UHP conditions were determined for garnet peridotites 123

from SBUC (Janák et al., 2006) and finally, diamond was found in kyanite and garnet-bearing 124

gneisses (Janák et al., 2015a) hosting UHP eclogites and garnet peridotites. The UHP 125

metamorphism in the Pohorje area resulted from subduction of the continental crust during 126

the Late Cretaceous time (Janák et al. 2004), documented by geochronological data (c.100-90 127

Ma; Thöni, 2002; Miller et al., 2005; Janák et al., 2009). 128

The studied eclogites are medium to coarse-grained, with macroscopically visible 129

reddish garnet, pale-green omphacite, bluish kyanite and pale zoisite, partly amphibolitised 130

with dark-green amphibole. Microscopic observations show that primary minerals – garnet, 131

omphacite, kyanite and phengite are variably replaced by retrograde ones forming 132

symplectites, i.e. diopside + plagioclase after omphacite, plagioclase + biotite after phengite 133

and sapphirine + corundum + spinel + anorthite after kyanite. Polycrystalline quartz 134

inclusions in garnet, omphacite and kyanite surrounded by radial cracks indicate breakdown 135

of former coesite. The most striking feature of omphacite is tiny needles and rods of quartz 136

(Fig. 2). They display a parallel orientation with the c-axis, indicating exsolution from a pre-137

existing, more silicic clinopyroxene. Geothermobarometry on the peak metamorphic 138

assemblage garnet + omphacite + kyanite + phengite + quartz/coesite records pressure and 139

temperature conditions of up to 3.5-3.7 GPa and 800-850 ºC (Vrabec et al., 2012). More 140

details on the kyanite eclogites from Pohorje can be found in Janák et al. (2004), Sassi et al. 141

(2004) or Vrabec et al. (2012), including unusually Cr-rich minerals in these eclogites - 142

kyanite, Mg-staurolite and corundum (Janák et al., 2015b). 143

The investigated samples come from 6 localities in Pohorje (Fig. 1): Jurišna Vas (JV4, 144

PO6), Novak (NO1), Tinjska Gora (PO4), Visole (VI04), Vranjekov Vrh (PO1) and Nova 145

Gora (NG1), see Vrabec et al. (2012) for a more detailed map and sample location. 146

Skogby et al.

7

147

Experimental methods 148

149

Infrared spectroscopy 150

All the seven omphacite samples were measured by Fourier transform infrared (FTIR) 151

absorption spectroscopy to characterize their OH vibrational bands. The spectrometer system 152

consisted of a Bruker Equinox 55 spectrometer equipped with a halogen lamp source, a CaF2 153

beam-splitter, a wire-grid polarizer (KRS-5) and an InSb detector. The crystals were oriented 154

by morphology and optical microscopy and polished on two sides parallel to (100) and (010). 155

To avoid cracks and turbid regions, the 100 to 200 m thick crystals sections were mounted 156

on circular 100-200 m apertures under the microscope. Polarized absorption spectra were 157

then acquired in the three principal optical directions (, and ) in the wavenumber range 158

2000-5000 cm-1

with a resolution of 4 cm-1

. 159

To obtain absorption areas for the OH bands the measured spectra were fitted using 160

the software Peakfit. A baseline subtraction was first applied before fitting with four Voight-161

shaped bands that produced a good fit to the experimental spectra. Water concentrations were 162

then calculated from the sum of the integrated absorption areas obtained in the , and 163

directions (Atot=A+A+A) according to Beer’s law using the calibrations of Koch-Müller et 164

al. (2007), Libowitzky & Rossman (1997) and Bell et al. (1995). 165

166

Mössbauer spectroscopy 167

The samples were measured by Mössbauer spectroscopy in order to determine the oxidation 168

state of iron, which is necessary for an accurate normalization of the structural formulae. 169

Since the samples contained ubiquitous inclusions and turbid areas, the point-source 170

technique (e.g. McCammon, 1994) was used. Small amounts (ca. 1-2 mg) of clear, inclusion-171

Skogby et al.

8

free crystal fragments were selected under the microscope. The samples were finely ground, 172

mixed with a thermoplastic resin and shaped to millimetre-sized cylindrical absorbers under 173

mild heating. Mössbauer spectra were acquired using a conventional spectrometer system 174

operated in constant-acceleration mode equipped with a 57

Co point source with a nominal 175

activity of 10 mCi and an active area of 0.5 x 0.5 mm. The absorbers were mounted on strip 176

tape and positioned close (< 1 mm) to the tip of the point source. Spectra were collected in 177

1024 channels over the velocity range -4.5 to 4.5 mms-1

, and were calibrated against an -iron 178

foil before folding and reduction to 256 channels. Spectral fitting was performed with the 179

least-squares program MDA (Jernberg & Sundqvist, 1983), using a fitting model with two 180

quadrupole doublets for Fe2+

and one doublet for Fe3+

. 181

182

Electron microprobe analysis 183

Electron microprobe analyses (EMPA) of the major elements (Al, Ti, Fe, Mg, Na, K, Si, Ca, 184

Mn, Cr) were performed at the Department of Earth Sciences, Uppsala University using a FE-185

EPMA JXA-8530F JEOL superprobe. Between 10 to 30 spot analyses were performed on 186

each sample using a beam current of 10 nA, an acceleration voltage of 15 kV and a beam 187

diameter of 1μm. A range of natural and synthetic compounds were used as standards: fayalite 188

(Fe), periclase (Mg), pyrophanite (Mn, Ti) corundum (Al), wollastonite (Ca and Si), eskolaite 189

(Cr), albite (Na) and orthoclase (K). 190

191

Results 192

193

The EMPA results show that the studied crystals are homogenous, whereas the compositions 194

for the different samples show substantial variation (Table 1). The samples have rather typical 195

omphacite compositions dominated by diopside (47-59 %) and jadeite (25-33 %) components, 196

Skogby et al.

9

with minor amounts also of hedenbergite (6-7 %), Ca-Tschermak (4-5 %), enstatite (3-4 %) 197

and CaEs (3-5 %) components. In back-scatter electron images, some samples show the 198

occurrence of micrometre-sized quartz-rods (Fig. 3), as previously described by Janák et al. 199

(2004) and Vrabec et al. (2012) in other crystals partly from the same samples. 200

Polarized FTIR spectra of oriented omphacite crystals show intense absorption bands 201

around 3470 and 3520 cm-1

, strongly polarized in the -direction (Fig. 4). Considerably 202

weaker bands polarized in the and -directions occur around 3620 cm-1

. For a few 203

samples (e.g. NO1), additional weak but sharp bands occur around 3670 cm-1

, indicative of 204

amphibole inclusions. These amphibole bands are stronger in spectra from regions close to 205

cracks and grain boundaries, and in turbid regions. By masking off the IR-beam by small 206

apertures (100-200 m) such regions could be avoided. Using the molar absorptivity of 207

65 000 L.mol

-1.cm

-2 established for omphacite by Koch-Müller et al., 2007, the recorded 208

spectra correspond to water concentrations in the range 530 – 870 ppm H2O (Table 1). Water 209

concentrations calculated based on the wavenumber-dependent general calibration of 210

Libowitzky & Rossman (1997) are in good agreement, with slightly higher values (up to 6 % 211

higher). However, the calibration by Bell et al. (1995) results in concentrations which are 212

approximately twice as high. Similar deviations were observed by Koch-Müller et al. (2004) 213

for a series of omphacites from Yakutian kimberlite pipes. The Bell et al. (1995) calibration 214

was based on an kimberlitic augite sample, which has considerably stronger bands at the 215

higher wavenumbers (3620 cm-1

) than typical omphacite spectra, and the observed deviation 216

is probably related to a wavenumber-dependence of the molar absorptivities. For the 217

evaluation of the experimental data we have used the mineral-specific calibration of Koch-218

Müller et al., 2007 which we consider to be that at present most accurate. 219

Mössbauer spectra were obtained by the point-source method for all samples, and 220

could be fitted satisfactorily by the three-doublet fitting model mentioned above. 221

Skogby et al.

10

Representative spectra are displayed in Fig. 5. The obtained hyperfine parameters are well in 222

line with previous studies of omphacite (e.g. Li et al., 2005), and are listed in Table 2, 223

together with assignments and absorption ratios. The results show that the Fe3+

/Fetot ratio vary 224

substantially, from 0.14 to 0.29 for the sample set. In line with Koch-Müller et al. (2007), the 225

two fitted Fe2+

doublets can be considered to be caused by Fe2+

in the M1 site (outer doublet) 226

and M2 site (inner doublet). However, the lack of resolution for these two doublets introduces 227

a large degree of uncertainty in the obtained Fe2+

distribution. In spite of this, the Fe3+

/Fetot 228

ratios are well constrained by the asymmetry of the two main high- and low-velocity 229

components in the measured spectra, as the Fe3+

contribution occurs entirely within the low-230

velocity component and are hence not affected by uncertainties in the Fe2+

distribution over 231

the M1 and M2 sites. 232

Structural formulae were calculated based on 12 positive charges taking all type of 233

cations into account, including the Fe2+

/Fe3+

distribution obtained from Mössbauer 234

spectroscopy and the hydrogen content obtained from FTIR spectroscopy (Table 1). Without 235

exception, this lead to cation sums less than 4 (excluding H+), which demonstrate the 236

occurrence of cation vacancies. CaEs (Ca0.5□0.5AlSi2O6) components were then calculated as 237

CaEs = 2 x vacancies per formula unit, and varied from 2.5 to 4.6 % for the studied samples. 238

239

Discussion 240

241

In evaluating intrinsic pyroxene water contents based on the intensity of OH-bands in IR-242

spectra, it is important to identify absorption bands from possible contaminants. These may be 243

caused by fluid or melt inclusions, as well as inclusions of hydrous minerals. Among hydrous 244

mineral inclusions, amphiboles have been shown to occur relatively frequently in 245

clinopyroxenes, particularly in samples from metamorphic rocks (e.g. Veblen & Buseck, 246

Skogby et al.

11

1981; Skogby et al., 1990). Fortunately, amphibole OH-bands normally have a characteristic 247

signature consisting of relatively sharp bands in the 3625-3725 cm-1

region (e.g. Skogby & 248

Rossman, 1991; Della Ventura et al., 2003), and can hence easily be distinguished from 249

intrinsic clinopyroxene OH-bands that occur at lower wavenumbers. More problematic, 250

however, is the possible occurrence of certain nanometer-sized sheet silicates. As 251

demonstrated by Koch-Müller et al. (2004) using TEM and synchrotron-IR, absorption bands 252

in the range 3600-3624 cm-1

of omphacite spectra can be caused by sub-micrometer 253

inclusions of clinochlore and amesite. In particular, this seems to be the case when strong 254

bands occur in this range (e.g. Katayama & Nakashima, 2003), and if not identified, may lead 255

to significant overestimation of the omphacite water contents. Nevertheless, weak bands that 256

normally occur in the 3620 cm-1

range of omphacite spectra (Skogby et al., 1990; Smyth et 257

al., 1991, Konzett et al., 2008a) may still be caused by intrinsic OH. This is not unexpected 258

since absorption bands in this range, with the same pleochroism (A A), are typical for 259

diopside as well as augite (e.g. Skogby, 2006) which both occur as significant components in 260

normal omphacite compositions. Furthermore, Bromiley & Keppler (2004) observed a major 261

band at 3613 cm-1

in spectra of synthetic jadeite and Na-rich clinopyroxene which they 262

interpreted to be caused by OH coupled to M2-site vacancies. In the present study, the 263

relatively weak bands in the 3620 cm-1

range were therefore included in the calculation of 264

omphacite water contents. The contribution from these bands was however minor, amounting 265

to 20-30 ppm H2O for the studied samples. 266

In general, the OH contents are expected to increase with the T, Ptot and PH2O 267

conditions, related to an increased amount of vacancies formed by stabilization of the CaEs 268

component. An increase in OH contents with equilibration pressure was also found by 269

Katayama & Nakashima (2003) for a series of clinopyroxenes from the Kokchetav UHP 270

metamorphic terrane. Experimental studies (e.g. Gasparik, 1986; Konzett et al., 2008b) have 271

Skogby et al.

12

shown that the CaEs component in omphacite in addition to bulk composition is also largely 272

controlled by temperature and pressure conditions. However, it should be noted that the OH 273

content of omphacite is not merely linked to equilibration pressure conditions. In their study 274

of omphacites from Yakutian kimberlite pipes, Koch-Müller et al. (2004) noted that samples 275

from the highest pressures environments (i.e. diamond-bearing eclogite xenoliths) showed the 276

lowest OH contents, which they interpreted to be caused by either low water activities during 277

crystallisation or, alternatively, to hydrogen loss during the uplift process. 278

The CaEs component, and hence the amount of vacancies at the structural M2 site, 279

display a positive correlation with the OH concentration in our samples (Fig. 6). This 280

observation is in line with previous studies (Smyth et al., 1991; Koch-Müller et al., 2004; 281

Katayama & Nikashima, 2003). Due to the determination of Fe2+

/Fe3+

ratios for the present 282

study, the underestimation in calculation of the CaEs component, sometimes resulting in 283

negative values, is here avoided. Although the regression line fitted to the experimental data 284

points in Fig. 6 is associated with some uncertainty, the intercept with the y-axis (i.e. for CaEs 285

= 0) indicates that an omphacite without the M2-vacancies may take up approximately 200 286

ppm H2O, and that the water content thereafter will increase progressively with the amount of 287

M2 vacancies. If the observed trend-line is extrapolated to the considerably higher CaEs 288

components of up to 10 mol-% indicated by the composition of primary omphacite estimated 289

from the modal content of the precipitated quartz rods mentioned above (Janák et al. 2004), 290

an original water content of ca 1500 ppm H2O is obtained. Due to the rather limited variation 291

in OH contents and CaEs components for the studied samples, this extrapolation is associated 292

with substantial uncertainty. However, the estimated content is in the similar range as that of 293

ca 2000 ppm H2O calculated by Smyth et al. (1991) for clinopyroxenes from South-African 294

kimberlite samples, based on a vacancy-rich precursor composition reconstructed from 295

exsolved garnet and kyanite. 296

Skogby et al.

13

In their study on hydroxyl contents in omphacite and omphacitic clinopyroxenes from 297

the Siberian platform, Koch-Müller et al. (2004) observed a pronounced correlation between 298

the absorbance of the OH-band at 3650-3540 cm-1

and the amount of Al in the tetrahedral site. 299

Such a correlation is not obvious for our sample series, however, the much more limited 300

variation in Al contents for our sample set (Table 1) obscures a possible trend. 301

The OH concentrations here obtained, ranging from 530-870 ppm H2O, are among the 302

higher reported for omphacite, and somewhat higher than those reported by Konzett et al. 303

(2008a) for three omphacite samples from Saualpe and Pohorje eclogites. Previous studies 304

have indeed reported even higher concentrations, but some of these were based on 305

calibrations that now have become obsolete (Skogby et al., 1990; Smyth et al., 1991) or on 306

FTIR spectra with OH bands that appear to be associated with hydrous phases (Katayama & 307

Nakashima, 2003). The high OH concentrations observed for the Pohorje omphacites are in 308

line with equilibration under hydrous UHP conditions. 309

310

Acknowledgements: Financial support from the Swedish Research Council (HS), the Slovak 311

Research and Development Agency (project APVV-0080-11to MJ) and the Slovak Scientific 312

Grant Agency VEGA (grant No. 2/0013/12 to MJ) is gratefully acknowledged. 313

References 314

Altherr, R., Lugovic, B., Meyer, H.P., Majer, V. (1995): Early Miocene post-collisional calc-315

alkaline magmatism along the easternmost segment of the periadriatic fault system 316

(Slovenia and Croatia). Mineral. Petrol., 54, 225–247. 317

Skogby et al.

14

Andrut, M., Wildner, M., Ingrin, J., Beran, A. (2007): Mechanisms of OH defect 318

incorporation in naturally occurring hydrothermally formed diopside and jadeite. Phys. 319

Chem. Minerals, 34, 543-549. 320

Bell, D.R., Ihinger P.H., Rossman G.R. (1995): Quantitative analysis of trace OH in garnet 321

and pyroxenes. Am. Mineral., 80, 465–474. 322

Bromiley, G.D. & Keppler, H. (2004): An experimental investigation of hydroxyl solubility in 323

jadeite and Na–rich clinopyroxenes. Contrib. Mineral. Petrol., 147, 189–200. 324

Bolfan-Casanova, N. (2005): Water in Earth’s mantle. Mineral. Mag., 69(3), 229-257. 325

De Hoog, J.C.M., Janák, M., Vrabec, M., Froitzheim, N. (2009): Serpentinised peridotites 326

from an ultra-high pressure terrane in the Pohorje Mts. (Eastern Alps, Slovenia): 327

geochemical constraints on petrogenesis and tectonic setting. Lithos, 109, 209–222. 328

Della Ventura, G., Hawthorne, F.C., Robert, J.-L., Iezzi, G. (2003): Synthesis and infrared 329

spectroscopy of amphiboles along the tremolite–pargasite join. Eur. J. Mineral., 15, 341–330

347. 331

Fodor, L.I., Gerdes, A., Dunkl, I., Koroknai, B., Pécskay, Z., Trajanova, M., Horváth, P., 332

Vrabec, M., Jelen, B., Balogh, K., Frisch, W. (2008): Miocene emplacement and rapid 333

cooling of the Pohorje pluton at the Alpine-Pannonian-Dinaridic junction, Slovenia. 334

Swiss J. Geosci., 101, 255–271. 335

Gasparik, T. (1986): Experimental study of subsolidus phase relations and mixing properties 336

of clinopyroxene in the silica-saturated system CaO-MgO-Al2O3-SiO2. Amer. Mineral., 337

71, 686-693. 338

Skogby et al.

15

Hinterlechner-Ravnik, A., Sassi, F.P., Visona, D. (1991): The Austridic eclogites, metabasites 339

and metaultrabasites from the Pohorje area (eastern Alps, Yugoslavia): 2. The 340

metabasites and metaultrabasites, and concluding considerations. Rend. Fis. Acc. Lincei, 341

2, 175–190. 342

Ingrin, J. & Skogby, H. (2000): Hydrogen in nominally anhydrous upper-mantle minerals: 343

concentration levels and implications. Eur. J. Mineral., 12, 543–570. 344

Janák, M., Cornell, D., Froitzheim, N., De Hoog, J.C.M., Broska, I., Vrabec, M., Hurai, V. 345

(2009): Eclogite-hosting metapelites from the Pohorje Mountains (Eastern Alps): P-T 346

evolution, zircon geochronology and tectonic implications. Eur. J. Mineral., 21, 1191–347

1212. 348

Janák, M., Froitzheim, N., Lupták, B., Vrabec, M., Krogh Ravna, E.J. (2004): First evidence 349

for ultrahigh-pressure metamorphism of eclogites in Pohorje, Slovenia: Tracing deep 350

continental subduction in the Eastern Alps. Tectonics. 23, TC5014. 351

doi:101029/2004TC001641. 352

Janák, M., Froitzheim, N., Vrabec, M., Krogh Ravna, E.J., Hoog, J.C.M. (2006): 353

Ultrahighpressure metamorphism and exhumation of garnet peridotite in Pohorje, Eastern 354

Alps. J. Metam. Geol., 24, 19–31. 355

Janák, M., Froitzheim, N., Yoshida, K., Sasinková, V., Nosko, M., Kobayashi, T., Hirajima, 356

T., Vrabec, M. (2015a): Diamond in metasedimentary crustal rocks from Pohorje, Eastern 357

Alps: a window to deep continental subduction. J. Metam. Geol., 33, 495-512, 358

doi:10.1111/jmg.12130. 359

Janák, M., Uher, P., Ravna, E.J.K., Kullerud, K., Vrabec, M. (2015b): Chromium-rich 360

kyanite, magnesiostaurolite and corundum in ultrahigh-pressure eclogites (examples from 361

Skogby et al.

16

Pohorje Mountains, Slovenia and Tromsø Nappe, Norway). Eur. J. Mineral., 27, 377-362

392. 363

Jernberg, P. & Sundqvist, T. (1983): A versatile Mössbauer analysis program. Uppsala 364

University, Institute of Physics (UUIP-1090). 365

Katayama, I. & Nakashima, S. (2003): Hydroxyl in clinopyroxene from the deep subducted 366

crust: evidence for H2O transport into the mantle. Amer. Mineral., 88, 229–234. 367

Katayama, I., Nakashima, S., Yurimoto, H. (2006): Water content in natural eclogite and 368

implication for water transport into the deep upper mantle. Lithos, 86, 245–259. 369

Kirst, F., Sandmann, S., Nagel, T.J., Froitzheim, N., Janák, M. (2010): Tectonic evolution of 370

the southeastern part of the Pohorje Mountains (Eastern Alps, Slovenia). Geol. Carpath., 371

61, 451–461. 372

Koch-Müller, M., Matsyuk, S.S., Wirth, R. (2004): Hydroxyl in omphacites and omphacitic 373

clinopyroxenes of upper mantle to lower crustal origin beneath the Siberian platform. 374

Amer. Mineral., 89, 921–931. 375

Koch-Müller, M., Abs-Wurmbach, I., Rhede, D., Kahlenberg, V., Matsyuk, S. (2007): 376

Dehydration experiments on natural omphacites: qualitative and quantitative 377

characterization by various spectroscopic methods. Phys. Chem. Minerals, 34, 663–678. 378

Konzett, J., Frost, D.J., Proyer, A., Ulmer, P. (2008b): The Ca-Eskola component in eclogitic 379

clinopyroxene as a function of pressure, temperature and bulk composition: an 380

experimental study to 15 GPa with possible implications for the formation of oriented 381

SiO2-inclusions in omphacite. Contrib. Mineral. Petrol., 155, 215–228. 382

Skogby et al.

17

Konzett, J., Libowitzky, E., Hejny, C., Miller, C., Zanetti, A. (2008a): Oriented quartz+calcic 383

amphibole inclusions in omphacite from the Saualpe and Pohorje Mountain eclogites, 384

Eastern Alps — An assessment of possible formation mechanisms based on IR- and 385

mineral chemical data and water storage in Eastern Alpine eclogites. Lithos, 106, 336-386

350. 387

Li, Y.L., Zheng, Y.F., Fu, B. (2005): Mössbauer spectroscopy of omphacite and garnet pairs 388

from eclogites; application to geothermobarometry. Amer. Mineral., 90, 90–100. 389

Libowitzky, E. & Rossman, G.R. (1997): An IR absorption calibration for water in minerals. 390

Amer Mineral, 82, 1111–1115. 391

McCammon, C.A. (1994): A Mössbauer milliprobe: Practical considerations. Hyperf. 392

Interact. 92, 1235-1239. 393

Miller, C., Mundil, R., Thöni, M., Konzett, J. (2005): Refining the timing of eclogite 394

metamorphism: a geochemical, petrological, Sm-Nd and U-Pb case study from the 395

Pohorje Mountains, Slovenia (Eastern Alps). Contrib. Mineral. Petrol., 150, 70–84. 396

Mioč, P. & Žnidarčič, M. (1977): Geological map of SFRJ 1:100 000, Sheet Slovenj Gradec. 397

Geological Survey, Ljubljana, Federal Geological Survey, Beograd. 398

O’Leary, J.A., Gaetani, G.A., Hauri, E.H. (2010): The effect of tetrahedral Al3+

on the 399

partitioning of water between clinopyroxene and silicate melt. Earth. Planet. Sci. Lett., 400

297, 111-120. 401

Peslier, A.H. (2010): A review of water contents of nominally anhydrous natural minerals in 402

the mantles of Earth, Mars and the Moon. J. Volcan. Geotherm. Res., 197, 239-258. 403

Skogby et al.

18

Sassi, R., Mazzolli, C., Miller, C., Konzett, J. (2004): Geochemistry and metamorphic 404

evolution of the Pohorje Mountain eclogites from the easternmost Austroalpine basement 405

of the Eastern Alps (Northern Slovenia). Lithos, 78, 235–261. 406

Skogby, H. (2006): Water in natural mantle minerals I: pyroxenes. Rev. Mineral. Geochem., 407

62, 155–168. 408

Skogby, H. & Rossman, G.R. (1991): The intensity of amphibole OH bands in the infrared 409

absorption spectrum. Phys. Chem. Mineral., 18, 64–68. 410

Skogby, H., Bell, D.R., Rossman, G.R. (1990): Hydroxide in pyroxene: variations in the 411

natural environment. Amer. Mineral., 75, 764–774. 412

Smyth, J.R., Bell, D.R., Rossman, G.R. (1991) Incorporation of hydroxyl in upper-mantle 413

clinopyroxenes. Nature, 351, 732–735. 414

Thöni, M.( 2002): Sm-Nd isotope systematics in garnet from different lithologies (Eastern 415

Alps): age results and an evaluation of potential problems for garnet Sm-Nd 416

chronometry. Chem. Geol., 185, 255–281. 417

Trajanova, M., Pécskay, Z., Itaya, T. (2008): K-Ar geochronology and petrography of the 418

Miocene Pohorje Mountains batholith (Slovenia). Geol. Carpath., 59, 247–260. 419

Uher, P., Janák, M., Konečný, P., Vrabec, M. (2014): Rare-element granitic pegmatite of 420

Miocene age emplaced in UHP rocks from Visole, Pohorje Mountains (Eastern Alps, 421

Slovenia): accessory minerals, monazite and uraninite chemical dating. Geol. Carpath., 422

65, 131–146. 423

Skogby et al.

19

Veblen, D.R. & Buseck, P.R. (1981): Hydrous pyriboles and sheet silicates in pyroxenes and 424

uralites: Intergrown microstructures and reaction mechanisms. Amer. Mineral., 66, 1107-425

1134. 426

Vrabec, M., Janák, M., Froitzheim, N., De Hoog, J.C.M. (2012): Phase relations during peak 427

metamorphism and decompression of the UHP kyanite eclogites, Pohorje Mountains 428

(Eastern Alps, Slovenia). Lithos, 144-145, 40-55. 429

Wade, J.A., Plank, T., Hauri, E.H., Kelley, K.A., Roggensack, K., Zimmer, M. (2008): 430

Prediction of magmatic water contents via measurement of H2O in clinopyroxene 431

phenocrysts, Geology, 36, 799-802. 432

Weis, F.A., Skogby, H., Troll, V.R., Deegan F.M., Dahren, B. (2015): Magmatic water 433

contents determined through clinopyroxene: Examples from the Western Canary Islands, 434

Spain. Geochem., Geophys., Geosys., 16, 2127–2146. 435

436

Skogby et al.

20

Title of tables 437

Table 1. Chemical composition of omphacite based on microprobe and FTIR analyses. 438

Table 2. Mössbauer hyperfine parameters for omphacite. 439

440

Skogby et al.

21

Table 1. Chemical composition of omphacite based on microprobe and FTIR analyses. 441

442

Fe valence ratios determined by Mössbauer spectroscopy. 443

444

445

NO1 PO6 JV4 PO4 VI04 NG1 PO1

Wt-% oxides SiO2 55.19 55.36 54.87 55.52 55.27 55.42 55.35 Al2O3 7.78 8.36 8.33 9.76 8.84 8.76 9.24 TiO2 0.13 0.13 0.12 0.13 0.12 0.13 0.12 Cr2O3 0.28 0.05 0.11 0.02 0.18 0.13 0.15 MgO 12.39 11.75 11.63 10.56 11.19 11.47 10.98 FeO 1.69 2.00 1.93 2.01 1.86 2.04 1.86 Fe2O3 0.47 0.60 0.65 0.79 0.83 0.38 0.69 MnO 0.02 0.03 0.02 0.07 0.03 0.02 0.02 CaO 18.91 18.28 18.08 16.53 17.50 18.01 17.45 Na2O 3.59 3.99 3.87 4.77 4.36 4.04 4.41 H2O 0.0535 0.0549 0.0636 0.0702 0.0769 0.0854 0.0866 Total 100.52 100.60 99.67 100.21 100.26 100.49 100.37 Structural formulae normalised to 12 charges Si 1.952 1.954 1.954 1.957 1.954 1.955 1.952 Al 0.324 0.348 0.349 0.406 0.368 0.364 0.384 Ti 0.004 0.004 0.004 0.004 0.004 0.004 0.004 Cr 0.008 0.001 0.003 0.001 0.005 0.004 0.004 Mg 0.654 0.618 0.617 0.555 0.589 0.603 0.578 Fe2+ 0.065 0.076 0.074 0.077 0.071 0.078 0.070 Fe3+ 0.016 0.020 0.022 0.027 0.028 0.013 0.024 Mn 0.001 0.001 0.001 0.003 0.001 0.001 0.001 Ca 0.717 0.691 0.690 0.624 0.663 0.681 0.659 Na 0.247 0.273 0.267 0.325 0.299 0.276 0.302 Total 3.986 3.987 3.981 3.977 3.982 3.978 3.978 H2O (ppm) 535 549 636 702 769 854 866 End-member components Di 0.586 0.549 0.543 0.474 0.515 0.537 0.510 Jd 0.247 0.274 0.268 0.327 0.300 0.278 0.303 Hd 0.058 0.068 0.065 0.065 0.062 0.069 0.062 CaTs 0.048 0.046 0.046 0.043 0.046 0.045 0.048 En 0.034 0.033 0.039 0.045 0.034 0.038 0.034 Ca-Esk 0.027 0.025 0.038 0.046 0.036 0.043 0.043

Skogby et al.

22

Table 2. Mössbauer hyperfine parameters for omphacite. 446

Sample Assignment cs dq fwhm Int 447

NO1 Fe2+ (1) 1.17 2.49 0.51 36.9 448

Fe2+ (2) 1.14 1.95 0.50 43.0 449

Fe3+ 0.41 0.43 0.55 20.1 450

451

PO6 Fe2+ (1) 1.16 2.56 0.50 30.6 452

Fe2+ (2) 1.13 2.01 0.51 48.2 453

Fe3+ 0.41 0.40 0.52 21.2 454

455

JV4 Fe2+ (1) 1.17 2.58 0.53 30.4 456

Fe2+ (2) 1.12 2.04 0.52 46.4 457

Fe3+ 0.42 0.44 0.51 23.3 458

459

PO4 Fe2+ (1) 1.14 2.59 0.52 36.3 460

Fe2+ (2) 1.17 1.94 0.51 37.6 461

Fe3+ 0.41 0.46 0.52 26.1 462

463

VI04 Fe2+ (1) 1.17 2.58 0.52 32.7 464

Fe2+ (2) 1.14 2.00 0.53 38.6 465

Fe3+ 0.40 0.42 0.62 28.7 466

467

NG1 Fe2+ (1) 1.16 2.61 0.50 31.4 468

Fe2+ (2) 1.15 1.97 0.51 54.2 469

Fe3+ 0.42 0.41 0.53 14.5 470

471

PO1 Fe2+ (1) 1.16 2.63 0.58 34.5 472

Fe2+ (2) 1.14 1.98 0.57 40.4 473

Fe3+ 0.39 0.47 0.59 25.1 474

cs=centroid shift, dq=quadrupole splitting, fwhm=full 475

width at half maximum, Int=percentage of total absorption area. 476

477

478

Skogby et al.

23

Figure captions 479

480

Fig. 1. Simplified geological map of Pohorje and adjacent areas (modified from Mioč & 481

Žnidarčič 1997), showing location of investigated rocks (white stars) near Slovenska Bistrica. 482

483

Fig. 2. Microphotograph of omphacite (Omp) with quartz (Qz) rods and inclusion surrounded 484

by radial cracks. Optical microscope, transmitted light, crossed polars. Sample PO6. 485

486

Fig. 3. Back-scatter electron image of omphacite (sample PO6) showing crystallographically 487

oriented quartz-rods. Length of arrow 20 m. 488

489

Fig. 4. Polarized FTIR spectra of omphacite normalized to 1 mm thickness. Spectra are 490

vertically off-set for clarity. Sinusoidal shortwave variations are interference fringes. Note 491

weak bands around 3670 cm-1

in the and directions for sample NO1 caused by amphibole 492

inclusions. 493

494

Fig. 5. Room-temperature Mössbauer spectra of omphacite with fitted sub-spectra. Blue and 495

red sub-spectra are assigned to Fe2+

, green sub-spectra are assigned to Fe3+

. 496

497

Fig. 6. Plot of omphacite water content as a function of Ca-Eskola (Ca0.5□0.5AlSi2O6) 498

component. 499

500

Skogby et al.

24

501

Fig. 1 502

Skogby et al.

25

503

Fig. 2 504

505

506

Fig. 3. 507

508

Skogby et al.

26

509

510

Fig. 4. 511

512

513

Skogby et al.

27

514

Fig. 5. 515

Skogby et al.

28

516

517

Fig. 6. 518

519